Sign up for our newsletter and get the latest HPC news and analysis.
Send me information from insideHPC:


The Use of High-Performance Polymers in HPC and Data Center Applications

“Polymer components in liquid cooling systems are attractive for several reasons: they are lightweight, typically less expensive than metal counterparts, and are impervious to corrosion that can render parts inoperable or introduce debris into flow paths. The challenges with many polymers used to date, however, are their abilities to handle high temperatures and physical stressors without deforming, cracking or creeping. These shortcomings become significant when leaks occur, leading to downtime or damage to equipment.”

Deep Learning on Summit Supercomputer Powers Insights for Nuclear Waste Remediation

A research collaboration between LBNL, PNNL, Brown University, and NVIDIA has achieved exaflop (half-precision) performance on the Summit supercomputer with a deep learning application used to model subsurface flow in the study of nuclear waste remediation. Their achievement, which will be presented during the “Deep Learning on Supercomputers” workshop at SC19, demonstrates the promise of physics-informed generative adversarial networks (GANs) for analyzing complex, large-scale science problems.

Intersect360 Research Examines Spending Trends in Machine Learning Market

Intersect360 Research has released a pair of new reports examining major technology trends in AI and machine learning, including the worldwide market, spending trends, and impact on HPC. “Machine learning has been in a very high growth stage,” says Intersect360 Research CEO Addison Snell. “In addition to that $10 billion, many systems not one hundred percent dedicated to machine learning are serving training needs as part of their total workloads, increasing the influence that machine learning has on spending and configuration.”

Data Center Transformation: Why a Workload-Driven and Scalable Architecture Matters

The QCT Platform on Demand (QCT POD) combines advanced technology with a unique user experience to help enterprises reach better performance and gain more insights. “With flexibility and scalability, QCT POD enables enterprises to address a broader range of HPC, Deep Learning, and Data Analytic demands that fulfill various applications.”

New ImmersionFAQ site to Dispel Misconceptions about Datacenter Immersive Cooling

Today GRC and Asperitas launched ImmersionFAQ.com, a new website with FAQs and facts about immersion cooling. “The newly developed vendor agnostic website is a key tool for operators looking to educate themselves on immersion and get answers to technical questions about liquid cooling.”

Google and NASA Achieve Quantum Supremacy

Today Google officially announced that it has achieved a major computing milestone. In partnership with NASA and Oak Ridge National Laboratory, the company has demonstrated the ability to compute in seconds what would take even the largest and most advanced supercomputers thousands of years, achieving a milestone known as quantum supremacy. “Achieving quantum supremacy means we’ve been able to do one thing faster, not everything faster,” said Eleanor Rieffel, co-author on the paper.

Deep Learning for Natural Language Processing – Choosing the Right GPU for the Job

In this new whitepaper from our friends over at Exxact Corporation we take a look at the important topic of deep learning for Natural Language Processing (NLP) and choosing the right GPU for the job. Focus is given to the latest developments in neural networks and deep learning systems, in particular a neural network architecture called transformers. Researchers have shown that transformer networks are particularly well suited for parallelization on GPU-based systems.

Report: Mellanox ConnectX Ethernet NICs Outperforming Competition

Today Mellanox announced that laboratory tests by The Tolly Group prove its ConnectX 25GE Ethernet adapter significantly outperforms the Broadcom NetXtreme E series adapter in terms of performance, scalability and efficiency. “Our testing shows that with RoCE, storage traffic, and DPDK, the Mellanox NIC outperformed the Broadcom NIC in throughput and efficient CPU utilization. ConnectX-5 also used ‘Zero-Touch RoCE’ to deliver high throughput even with partial and no congestion control, two scenarios where Broadcom declined to be tested.”

Podcast: Building AI Datacenters with NVIDIA DGX Reference Architectures

In this podcast, Lee Carter from Bright Computing and Carlo Ruiz from NVIDIA discuss building datacenters with NVIDIA DGX Reference Architectures. “NVIDIA DGX servers managed by Bright Cluster Manager deliver the perfect balance of increased concurrency for handling data science workloads, the massive compute requirements associated with those workloads, and the seamless management of all resources across the HPC cluster. Join this podcast to learn more about the power behind this partnership.”

Study finds On-Premise Systems Critical to Enterprise AI

Today Cray released the findings of its industry report on the State of Enterprise AI Adoption 2019, which reveals that performance, data locality and security are the top considerations for choosing an AI infrastructure. “The survey assessed whether organizations and individuals are prepared to run AI effectively and to take advantage of the capabilities it offers.”